Tobias Schreieder | Ph.D. Student at TU Dresden & ScaDS.AI

About Me

I am a Ph.D. student at TU Dresden and ScaDS.AI. My research interests lie in the areas of Natural Language Processing, Information Retrieval, Trustworthy AI and AI4Science. With a focus on evidence-based text generation with LLMs, I develop methods that allow users to trace LLM-generated content back to their underlying sources through citations.

Research Interests

Evidence-based Text Generation with LLMs: Developing methods that enable large language models to generate traceable and verifiable responses by grounding their outputs in supporting evidence.
Evidence Retrieval & Fact Verification: Building retrieval and reasoning systems that identify scientific evidence for factual claims and improve the verification of multilingual scientific content.
AI4Science: Applying large language models to support scientific discovery by assessing the novelty of research ideas and assisting researchers in the scientific discovery process.
Argument Mining & Stance Detection: Investigating how computational methods can identify argumentative structures and detect supporting or opposing viewpoints in text and images.
Smartwatch Privacy: Studying privacy risks in wearable sensor data and developing methods to protect sensitive health information against re-identification attacks while preserving data utility.

News

[Jul. 2026] I presented our survey paper Attribution, Citation, and Quotation: A Survey of Evidence-based Text Generation with Large Language Models at ACL 2026 in San Diego, USA.
[Jun. 2026] Two papers were accepted at CLEF 2026: Scientific Claim-Source Retrieval Revisited: A Comparative Study of Style Transfer and Re-Ranking, following our selection for Best of Labs 2025, and Claim2Source at CheckThat! 2026: Improving Multilingual Scientific Claim-Source Retrieval with Verification-based Re-Ranking. Our Claim2Source system ranked 1st in the CheckThat! 2026 Task 1 on Source Retrieval for Scientific Web Claims.
[Jun. 2026] I gave an invited talk on Evidence-based Text Generation with Large Language Models at the ScaDS.AI Summer School 2026 in Leipzig and presented our demo system SQuAI: Exploring Scientific Knowledge with AI at OUTPUT.DD and the Dresden Science Night.
[Jan. 2026] Our research project EVIDENZ: Evidence-Based Text Generation with Large Language Models started. I lead the project in collaboration with Springer Nature. The project runs from 2026–2027 and is funded with €115,000 through the Software Campus program funded by BMFTR.
[Oct. 2025] Our paper SQuAI: Scientific Question-Answering with Multi-Agent Retrieval-Augmented Generation was published at CIKM 2025.